[BUG][RT][ti-j72xx] spi: cadence-quadspi: boot failed with rtmutex deadlock detected

[BUG][RT][ti-j72xx] spi: cadence-quadspi: boot failed with rtmutex deadlock detected

[He, Guocai (CN)]

Hi all,

I am seeing an reproducible issue in
drivers/spi/spi-cadence-quadspi.c on TI J72xx platform.

Kernel:
  v6.1/standard/preempt-rt/ti-sdk-6.1/ti-j7xxx
  (based on Linux v6.1.159 PREEMPT_RT)

Platform:
  TI J72xx
  QSPI controller: Cadence QuadSPI

Problem description:
when boot fail with the following call trace:
[    2.028205] ------------[ cut here ]------------
[    2.028213] rtmutex deadlock detected
[    2.028227] WARNING: CPU: 1 PID: 50 at kernel/locking/rtmutex.c:1642 __rt_mutex_slowlock_locked.constprop.0+0x184/0x2dc
[    2.028246] Modules linked in:
[    2.028256] CPU: 1 PID: 50 Comm: kworker/u8:1 Not tainted 6.1.159-rt32-yocto-preempt-rt #1
[    2.028263] Hardware name: Texas Instruments AM62A7 SK (DT)
[    2.028269] Workqueue: events_unbound deferred_probe_work_func
[    2.028282] pstate: 60000005 (nZCv daif -PAN -UAO -TCO -DIT -SSBS BTYPE=--)
[    2.028288] pc : __rt_mutex_slowlock_locked.constprop.0+0x184/0x2dc
[    2.028294] lr : __rt_mutex_slowlock_locked.constprop.0+0x184/0x2dc
[    2.028300] sp : fffffe000abef090
[    2.028302] x29: fffffe000abef110 x28: fffffe00090dc620 x27: 0000000000000001
[    2.028311] x26: fffffc0801af98f8 x25: 0000000000000000 x24: fffffc0804ac4088
[    2.028319] x23: 0000000000000000 x22: fffffe000abef098 x21: 00000000ffffffdd
[    2.028327] x20: fffffc0801a01d80 x19: fffffc08034b53e0 x18: ffffffffffffffff
[    2.028335] x17: 0000000000000000 x16: 0000000000000001 x15: fffffe000abef5b0
[    2.028343] x14: fffffc0804d1ad8e x13: 0a64657463657465 x12: 64206b636f6c6461
[    2.028351] x11: fffffffffffe0000 x10: 2d2d2d2d2d2d2d2d x9 : fffffe00080fc87c
[    2.028359] x8 : fffffe000abef090 x7 : 6b636f6c64616564 x6 : 000000000000000c
[    2.028367] x5 : 00000000fffff141 x4 : fffffe000abeee40 x3 : 0000000000000000
[    2.028375] x2 : 0000000000000000 x1 : 0000000000000000 x0 : fffffc0801a01d80
[    2.028384] Call trace:
[    2.028386]  __rt_mutex_slowlock_locked.constprop.0+0x184/0x2dc
[    2.028393]  mutex_lock+0x98/0xbc
[    2.028401]  spi_controller_resume+0x28/0xbc
[    2.028410]  cqspi_resume+0x58/0xa4
[    2.028418]  pm_generic_runtime_resume+0x34/0x50
[    2.028426]  __genpd_runtime_resume+0x38/0x90
[    2.028434]  genpd_runtime_resume+0x148/0x27c
[    2.028443]  __rpm_callback+0x50/0x1b4
[    2.028452]  rpm_callback+0x74/0x80
[    2.028459]  rpm_resume+0x5e4/0x85c
[    2.028467]  __pm_runtime_resume+0x5c/0xc0
[    2.028474]  cqspi_exec_mem_op+0x64/0xf50
[    2.028481]  spi_mem_exec_op+0x3bc/0x410
[    2.028487]  spinand_reset_op.constprop.0+0x64/0xa0
[    2.028495]  spinand_probe+0xdc/0x650
[    2.028502]  spi_mem_probe+0x74/0xc4
[    2.028508]  spi_probe+0x8c/0xec
[    2.028515]  really_probe+0xc4/0x2e0
[    2.028523]  __driver_probe_device+0x80/0x120
[    2.028531]  driver_probe_device+0xdc/0x170
[    2.028539]  __device_attach_driver+0xc0/0x14c
[    2.028548]  bus_for_each_drv+0x88/0xec
[    2.028555]  __device_attach+0xa8/0x1ac
[    2.028563]  device_initial_probe+0x1c/0x30
[    2.028572]  bus_probe_device+0xa4/0xb0
[    2.028579]  device_add+0x380/0x784
[    2.028586]  __spi_add_device+0x74/0x114
[    2.028592]  spi_add_device+0x68/0xa0
[    2.028597]  spi_register_controller+0x730/0xb00
[    2.028602]  cqspi_probe+0x690/0x98c
[    2.028608]  platform_probe+0x70/0xcc
[    2.028615]  really_probe+0xc4/0x2e0
[    2.028623]  __driver_probe_device+0x80/0x120
[    2.028631]  driver_probe_device+0xdc/0x170
[    2.028639]  __device_attach_driver+0xc0/0x14c
[    2.028648]  bus_for_each_drv+0x88/0xec
[    2.028655]  __device_attach+0xa8/0x1ac
[    2.028663]  device_initial_probe+0x1c/0x30
[    2.028671]  bus_probe_device+0xa4/0xb0
[    2.028679]  deferred_probe_work_func+0x90/0xd0
[    2.028687]  process_one_work+0x1f8/0x4b0
[    2.028696]  worker_thread+0x238/0x440
[    2.028703]  kthread+0x138/0x140
[    2.028710]  ret_from_fork+0x10/0x20
[    2.028718] ---[ end trace 0000000000000000 ]---

Analysis:
* 18db8c7f20a6 (HEAD) Revert "spi: cadence-qspi: fix pointer reference in runtime PM hooks"
*   7083c6620486 Merge branch 'v6.1/standard/base' into v6.1/standard/preempt-rt/ti-sdk-6.1/ti-j7xxx
|\
| *   2c1b513331dc Merge tag 'v6.1.140' into v6.1/standard/base
| |\
| | * da3c5173c55f (tag: v6.1.140) Linux 6.1.140
| | * 05e85d376720 x86/modules: Set VM_FLUSH_RESET_PERMS in module_alloc()

It is ok at da3c5173c55f，but it is bad at 7083c6620486.
I revert the commit(2c914aac952) which merged from stable release,  it also ok at 18db8c7f20a6.
Maybe it is a compatibility issue between v6.1/standard/preempt-rt/ti-sdk-6.1/ti-j7xxx and the stable release.

if this is an known issue ? 

Thanks,
Guocai He

Re: [BUG][RT][ti-j72xx] spi: cadence-quadspi: boot failed with rtmutex deadlock detected

[Mark Brown]

[-- Attachment #1: Type: text/plain, Size: 1192 bytes --]

On Wed, Dec 17, 2025 at 08:21:19AM +0000, He, Guocai (CN) wrote:

> Kernel:
>   v6.1/standard/preempt-rt/ti-sdk-6.1/ti-j7xxx
>   (based on Linux v6.1.159 PREEMPT_RT)

v6.1 is pretty old.

> It is ok at da3c5173c55f，but it is bad at 7083c6620486.
> I revert the commit(2c914aac952) which merged from stable release,  it also ok at 18db8c7f20a6.
> Maybe it is a compatibility issue between v6.1/standard/preempt-rt/ti-sdk-6.1/ti-j7xxx and the stable release.

> if this is an known issue ? 

This isn't something I've heard about before, and there are no fixes for
that commit in -next.  It's entirely plausible that whatever issue this
is also exists in mainline, looking at the patch it's likely there were
problems before this with the locking just not happening properly at all.

Please include human readable descriptions of things like commits and
issues being discussed in e-mail in your mails, this makes them much
easier for humans to read especially when they have no internet access.
I do frequently catch up on my mail on flights or while otherwise
travelling so this is even more pressing for me than just being about
making things a bit easier to read.

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 488 bytes --]

Re: [BUG][RT][ti-j72xx] spi: cadence-quadspi: boot failed with rtmutex deadlock detected

[Nishanth Menon]

On 08:21-20251217, He, Guocai (CN) wrote:
> Hi all,
> 
> I am seeing an reproducible issue in
> drivers/spi/spi-cadence-quadspi.c on TI J72xx platform.
> 
> Kernel:
>   v6.1/standard/preempt-rt/ti-sdk-6.1/ti-j7xxx
>   (based on Linux v6.1.159 PREEMPT_RT)

^^ sounds like TI vendor kernel?

> 
> Platform:
>   TI J72xx
>   QSPI controller: Cadence QuadSPI
> 

Will be good to discuss vendor kernel issues in e2e.ti.com please.
unless you can reproduce this issue on upstream stable or mainline, this
might be the wrong forum for discussion.

-- 
Regards,
Nishanth Menon
Key (0xDDB5849D1736249D) / Fingerprint: F8A2 8693 54EB 8232 17A3  1A34 DDB5 849D 1736 249D
https://ti.com/opensource